| Summary: | 碩士 === 國立成功大學 === 醫學工程研究所碩博士班 === 96 === Owing to the influence of the later marriage in recent years, the occurrence of Infertility is increasing on account to the aging of female ovarian function. Follicular number and growth are important factors for estimating Polycystic Ovary Syndrome (PCOS) and Assisted Reproductive Techniques (ART). In the past, two-dimensional ultrasound system has popularly been applied in clinical diagnosis to acquire the cross-sectional image of ovarian follicles. The development of Intravaginal ultrasound can also provide the information of three-dimension image further. Although we can utilize the imaging system to analyze 2D image information and reconstruct 3D model for single follicle, it is time-consuming and only obtains inaccurate information to analyze all follicles in the ovary. Thus, the first objective of this study was to develop an automatic boundary search method for replacing the manual one and further build multiple 3D models. Hopefully the objective measurement tool enables the clinicians in searching follicular boundary of ovary and provides more accurate information about follicles.
In the present study, we reconstructed the ultrasonic images and assessed the parameters of true follicles with the quantitative assessment system. In order to verify the accuracy of the systematic reconstruction method, we constructed a phantom model to imitate the ovarian follicles. The final results showed that our implementation could elevate the accuracy rate up to 90%. Thus, our results demonstrate that the proposed automatic algorithm was more efficient in follicular boundary detection compared to the manual one. The multiple reconstructed 3-D model could delineate the number, size and location of ovarian follicles clearly which can provide clinicians alternative tools and useful information for comprehend important ovarian diseases like polycystic ovary syndrome.
|
|---|
